Incubation of carbohydrate-free human serum albumin (HSA) with fructose in an aqueous buffer at pH 7.4 resulted in glycation of epsilon-amino groups of lysyl residues. A recently developed procedure, involving analysis of hexitol amino acids by high-performance liquid chromatography of phenylthiocarbamyl derivatives, was used to show that 85% of the bound hexose was attached to protein via carbon 2 (C-2). The remainder was attached to protein via carbon 1 (C-1). When incubations were conducted with glucose under identical conditions, all the hexose was attached via C-1. Examination of human ocular lens proteins showed that the majority of the covalently bound hexose was connected to epsilon-amino groups of lysyl residues via C-1; this was attributed mainly to nonenzymatic glucosylation in vivo, which has already been documented. A significant proportion (10-20%) of the bound hexose was connected via C-2. In view of the HSA-hexose incubation results (above), this indicated that the lens proteins had reacted with endogenous fructose; i.e., they had undergone nonenzymatic fructosylation in vivo. The model protein bovine pancreatic ribonuclease A reacted with fructose and glucose at similar rates under physiological conditions. However, covalent, non-disulfide cross-linking, which could be inhibited by D-penicillamine, was induced 10 times more rapidly by fructose than by glucose. It is postulated that some of the protein cross-linking that occurs in vivo is fructose-induced. The possible significance of these processes in diabetic subjects is discussed.

The site specificity of in vitro glycation of horse liver alcohol dehydrogenase (ADH) was examined and the results interpreted in terms of structural features of the enzyme molecule. In a phosphate buffer solution, glycation occurred at Lys231 (the main site of glycation in vivo), at Lys228 (which is not glycated in vivo), and at several unidentified positions. Buffer anions or NAD+ did not affect glycation of Lys231; this supported our hypothesis that the base catalyst which removes a proton from carbon 2 of a Lys231-attached aldimine is part of the ADH molecule [Shilton, B.H. & Walton, D.J. (1991) J. Biol. Chem. 266, 5587-5592]. Use of a molecular modelling programme indicated that this catalyst was likely to be the imidazole group of His348, exerting its effect through the hydroxyl of Thr347. Glycation of Lys228 occurred only in the presence of phosphate; in this case molecular modelling showed that the base catalyst could be a phosphate ion, bound to ADH at a positive region of the coenzyme binding site. NAD+ inhibited glycation of Lys228 by binding to the enzyme and restricting access to glucose.

No abstract

No abstract

No abstract